Can Progressive Vouchers Help the Poor Benefit from School Choice? Evidence from the Chilean Voucher System

This paper considers a major educational reform implemented in Chile, one that increased voucher payments by 50% for students in the lowest 40% of the income distribution. This increased the revenues that schools received for serving low-income students and lowered the relative prices of private voucher schools for eligible parents. I use a previously unavailable national dataset to implement a regression discontinuity design exploiting the fact that eligibility for the targeted vouchers is a discontinuous function of a socioeconomic ranking. The estimates allow me to reject that access to the higher voucher improved students' test scores by more than 0.04 standard deviations in math or language. This may reflect parents do not seem to have used the vouchers to choose schools with higher test scores or average socioeconomic status. My findings cast doubt on the potential of progressive vouchers to decrease the cream-skimming and improve the educational outcomes of the poor